Cognitive Networks in the Presence of I/Q Imbalance and Imperfect Csi: Receiver Design and Performance Analysis

Abstract

Future wireless communication systems, including fifth-generation (5G) networks and the Internet of Things (IoT), require a massive number of inexpensive transceivers. These transceivers come with various hardware impairments, such as phase noise and in-phase/quadrature phase (I/Q) imbalance. This piece of work studies the performance of underlay cognitive radio (CR) networks, considering the joint effect of I/Q imbalance and imperfect channel-state information (CSI) at the secondary user. In order to mitigate the effect of I/Q imbalance, an optimal maximum likelihood (ML) receiver design is proposed and analyzed. Specifically, a closed-form expression of the average pairwise error probability (APEP) and a tight upper bound of the average bit error rate (ABER) are derived. In addition, a widely linear equalization (WLE) receiver that has performance close to the optimal receiver with a computational complexity close to the traditional blind receiver is proposed. In particular, the exact PEP of this WLE receiver is obtained and its APEP is calculated numerically. Moreover, an exact expression is derived for Cramer-Rao lower bound (CRLB) of the secondary system receiver channel estimation error in the presence of I/Q imbalance at the secondary transmitter/receiver (STx/SRx) sides. Computer simulations prove the analytical results of the proposed receivers. The obtained results show that the optimal receiver has the best performance and the WLE receiver outperforms the traditional ML receiver in most cases. In addition, the analysis shows that the best estimator that reaches the CRLB is not affected by the I/Q imbalance at STx/SRx.